Competition year :
2020-2021
Deadline (pre-request) :
June 1st, 2020 at 11:00 (EST)
Deadline (application) :
July 2nd, 2020 at 11:00 (EST)
Announcement of results :
Mid -October 2020
Amount :
Maximum of CAD $100 000 per year* see details
Duration :
Maximum of 3 years, non renewable
CHIST-ERA is a consortium of research funding organisations in Europe and beyond supporting use- inspired basic research in Information and Communication Technologies (ICT) or at the interface between ICT and other domains. The CHIST-ERA consortium is itself supported by the European Union’s Future & Emerging Technologies (FET) programme. CHIST-ERA promotes novel and multidisciplinary research with the potential to lead to significant technology breakthroughs in the long term. The funding organisations jointly support high risk and high impact research projects selected in the framework of CHIST-ERA, in order to reinforce European capabilities in promising emerging topics.
Content of the Call
Topics: | Explainable Machine Learning-based Artificial Intelligence Novel Computational Approaches for Environmental Sustainability |
Indicative budget: | Approx. 16 M€ |
International consortium: | The project consortia must have a minimum of 3 eligible and independent partners requesting funding in at least 3 of the following countries including at least 2 partners from EU Member States or Associated Countries not taking into account UK: Austria, Belgium, Bulgaria, Québec (Canada), Czech Republic, Estonia, Finland, France, Greece, Hungary, Ireland, Israel, Italy, Latvia, Lithuania, Poland, Portugal, Romania, Slovakia, Spain, Sweden (topic 1 only), Switzerland, Turkey, United Kingdom (topic 1 only) |
Standard consortium size: | Three to six partners |
Evaluation: | Proposals are evaluated based on criteria of Relevance to the topic (short proposals only), Scientific and technological quality, Implementation and Impact |
Funding: | Each partner is funded separately by the national/regional funding organisation they are applying to. They must fulfil the conditions of their funding organisation, as described in the annex |
Tentative Timeline
14 February 2020, 17:00 CET | Deadline for short proposal submission |
End of April 2020 | Notification of accepted short proposals |
June 2020 | Deadline for full proposal submission |
Mid-October 2020 | Notification of accepted proposals |
1 December 2020 | First possible start date for accepted projects |
Research Targeted in the Call
The CHIST-ERA (www.chistera.eu) consortium has created a common funding instrument to support European research projects that engage in long-term research in the area of ICT and ICT-based sciences. Through this instrument, the national/regional funding organisations of CHIST-ERA support and join the Horizon 2020 Future and Emerging Technologies (FET) agenda. By launching joint European calls, they can support more diverse research communities, who are able to tackle the most challenging and novel research topics.
Each year, CHIST-ERA launches a call for research projects in two new topics of emergent scientific importance.
In previous years, CHIST-ERA calls have targeted quantum computing, consciousness, knowledge extraction, low-power computing, intelligent user interfaces, smart communication networks, adaptive machines, distributed computing, trustworthy cyber-physical systems, human language understanding, security and privacy in the IoT, terahertz communication, lifelong learning for intelligent systems, visual analytics, object recognition and manipulation by robots, big data and process modelling for smart industry, analog computing for artificial Intelligence and smart computing in networks.
This year’s call concerns the following topics:
- Explainable Machine Learning-based Artificial Intelligence (XAI);
- Novel Computational Approaches for Environmental Sustainability (CES).
A workshop was held in Tallinn (Estonia) on 11-13 June 2019, bringing together researchers from across a range of research communities and countries, to identify the challenges and promising research directions within the two selected topics. This open consultation has formed the scope of this call.
CHIST-ERA projects should be of a FET-like nature and contribute to the development of the European research and innovation capacity in the technology domain of the call topics. The transformative research done in CHIST-ERA should explore collaborative advanced interdisciplinary science and/or cutting-edge engineering with the potential to initiate or foster new lines of technology and help Europe grasp leadership early on in promising future ICT and ICT-based areas with potential for significant impact in the long term.
Open access to publications and research data is a key asset to leverage on research funding. Applicants are encouraged to consider approaches promoting open access starting from the project preparation stage (see p. 8 about CHIST-ERA developing policy and ongoing activities).
To widen participation throughout Europe, CHIST-ERA projects are encouraged to include partners from the so-called Widening Countries participating in the call: Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Portugal, Romania, Slovakia and Turkey.
To build leading innovation capacity across Europe and connect with industry, CHIST-ERA projects are encouraged to involve key actors that can make a difference in the future, for example excellent young researchers, ambitious high-tech SMEs etc.
1st Topic: Explainable Machine Learning-based Artificial Intelligence (XAI)
Explanation of decisions made by Artificial Intelligence (AI) systems is seen as important for the trust and social acceptance of AI. It is likely in the future that there will be a ‘right to an explanation’ for decisions that affect an individual. The objective of research on this topic is to make machine learning- based AI explainable.
To do this effectively, it is expected that explanation will need to be designed and integrated into AI systems from the outset, including the data collection and training of algorithms that are the basis of machine learning-based AI.
Along with the technical challenges, it is important to consider that explanation is required at different levels for different stakeholders with different levels of technical knowledge, and in different application domains. It is also important to measure the effectiveness of the explanation at the human and the technical levels, for example by evaluating how transparency, trust and usability are enhanced.
Target Outcomes
- Integration of explainability into new and existing AI systems, including:
- Explainability for identification and elimination of biases in data collection
- Explainability in the training of machine learning algorithms
- Development of algorithms and user interfaces for explainability
- Integration of social and ethical aspects of explainability into AI systems including: User requirements, bias, objectivity and trust
- Developing a means to measure the effectiveness of explainable systems for different stakeholders (objective benchmarks and evaluation strategies for research in this domain)
Applicants should also consider the following:
- Give due consideration to performance evaluation and experiment reproducibility
- The benefits of international collaboration
- Co-creation of projects with stakeholders, including end users, policy makers and industry
- Potential for development of standards or frameworks
- Responsible research and innovation including: Use and protection of data; The legal and ethical issues of providing explanations (what level of explanation is required or appropriate for whom); Open access to research data and publications
Expected Impacts
- Development of novel, ambitious and reliable technologies for the different components of explainable machine learning-based AI, including: AI systems with integrated explanations in a variety of application areas; Frameworks for integrating explainability into AI (Explainability by Design); Methods for putting explainability into current AI systems; Use cases in specific application areas
- Identification of new opportunities and applications fostered through explainable AI
- Enhanced interdisciplinarity; Stakeholders involvement in design and implementation of explainable AI systems; Consideration of the ethical and social aspects of explainability in AI systems
- Widened participation throughout Europe by involving partners from the Widening Countries
- Reinforced innovation capacity across Europe by involvement of key actors, for example young researchers, high-tech SMEs or first-time participants
2nd Topic: Novel Computational Approaches for Environmental Sustainability (CES)
With the challenge of environmental changes being highlighted, it is important that scientists are able to understand and model the environment so they can understand and predict upcoming changes. As environmental models become more complex and more adaptable in real time, it is necessary to change the way we work with these models, to be more integrative, more reactive and reduce the amount of computational power being used. This will improve the computational models that we have and allow better predictions on the future of our planet.
Better data : Better model : Better prediction : Better decision/action
Target Outcomes
- Improvements to computational systems so that data be collected and modelled
- In real time
- At different levels of complexity and granularity
- Integration of models to improve overall knowledge of an area or system
- Displaying the outputs of a model in a way that different stakeholders are able to understand and make decisions from them
- Modelling of uncertainty in a way that is easy to understand and make decisions from
Applicants should also consider the following:
- Cross traditional boundaries between disciplines in order to strengthen the communities involved in tackling these new challenges
- The benefits of international collaboration
- Co-creation of projects with stakeholders, including end users, policy makers and industry
- Potential for development of standards or frameworks
- Responsible research and innovation including: Use and protection of modelling data; How to reduce the environmental impact of the computational power used for modelling; Open access to data, models and publications
Expected Impacts
- Novel and ambitiously improved methods for environmental modelling, including whole systems approaches; Increased integration of models and data; Increased standardisation of environmental data approaches and storage
- Improved tools for displaying the outputs of the modelling, including the uncertainty in the system; Effective usage of these tools by stakeholders and policy makers using these tools
- Enhanced interdisciplinarity; Stakeholders involvement in research projects design and implementation
- Widened participation throughout Europe by involving partners from the Widening Countries
- Reinforced innovation capacity across Europe by involvement of key actors, for example young researchers, high-tech SMEs or first-time participants